Blockchain-Based Federated Learning With Enhanced Privacy and Security Using Homomorphic Encryption and Reputation

Federated learning, leveraging distributed data from multiple nodes to train a common model, allows for the use of more data to improve the model while also protecting the privacy of original data. However, challenges still exist in ensuring privacy and security within the interactions. To address these issues, this article proposes a federated learning approach that incorporates blockchain, homomorphic encryption, and reputation. Using homomorphic encryption, edge nodes possessing local data can complete the training of ciphertext models, with their contributions to the aggregation being evaluated by a reputation mechanism. Both models and reputations are documented and verified on the blockchain through the consensus process, which then determines the rewards based on the incentive mechanism. This approach not only incentivizes participation in training, but also ensures the privacy of data and models through encryption. Additionally, it addresses security risks associated with both data and network attacks, ultimately leading to a highly accurate trained model. To enhance the efficiency of learning and the performance of the model, a joint adaptive aggregation and resource optimization algorithm is introduced. Finally, simulations and analyses demonstrate that the proposed scheme enhances learning accuracy while maintaining privacy and security.

I. Introduction

In recent years, machine learning has been extensively applied in Industrial Internet of Things (IIoT), playing a crucial role in processing and making decisions based on industrial data [1], [2]. As the volume of industrial equipment and data has grown rapidly, there has been a paradigm shift in machine learning from centralized to distributed training. However, industrial data often encompasses valuable and sensitive information. This sensitivity poses a significant challenge for factories when it comes to data sharing and collaboration, leading to an inability to fully harness the potential value of the data [3]. Federated learning, a notable distributed framework, facilitates local learning on distributed data and procures an effective global learning model through iterative aggregation of parameters. This method not only ensures data privacy but also significantly reduces communication [4], [5], [6]. Concurrently, edge computing, which relocates computing and storage resources from the center to the network edges, offers robust distributed service with real-time response [7]. As a result, federated learning in edge computing has emerged as a practical solution. However, it is worth noting that vulnerabilities and security risks within distributed nodes and their communications present significant challenges [8], [9].